High-throughput Protein and mRNA Expression–based Classification of Gastric Cancers Can Identify Clinically Distinct Subtypes, Concordant With Recent Molecular Classifications

Gastric cancers have recently been classified into several types on the basis of molecular characterization, and the new taxonomy has shown to have clinical relevance. However, the technology required for thorough molecular classification is complicated and expensive, currently preventing widespread use. We aimed to reproduce the results of molecular classification using only simple techniques, that is, immunohistochemical analysis and in situ hybridization. We classified a cohort of 349 successive gastric adenocarcinomas into 5 subtypes, on the basis of protein or mRNA expression of MLH1, E-cadherin, p53, and Epstein-Barr virus. We observed that the subtypes presented distinct clinicopathologic characteristics and corresponded to the molecular classifications previously reported. Epstein-Barr virus -positive tumors were more common in male individuals and in the body of the stomach. Microsatellite-unstable (MSI) tumors, which showed aberrant MLH1 expression, were correlated with increased age and intestinal histology. Both types showed better overall survival than the other types. Gastric cancers with reduced expression of E-cadherin, corresponding to the epithelial to mesenchymal transition or genome stable subtypes, showed the poorest overall survival, with a high prevalence of poorly cohesive carcinoma (ie, diffuse type, of the Lauren classification system). In conclusion, we were able to reproduce a previously reported molecular classification of gastric cancers using immunohistochemical analysis and in situ hybridization. We verified the effectiveness and applicability of this method, which shows promise for use in a clinical setting in the foreseeable future.